Neuroblastoma and pre-B lymphoma cells share expression of key transcription factors but display tissue restricted target gene expression
- 5.7k Downloads
Transcription factors are frequently involved in the process of cellular transformation, and many malignancies are characterized by a distinct genetic event affecting a specific transcription factor. This probably reflects a tissue specific ability of transcription factors to contribute to the generation of cancer but very little is known about the precise mechanisms that governs these restricted effects.
To investigate this selectivity in target gene activation we compared the overall gene expression patterns by micro-array analysis and expression of target genes for the transcription factor EBF in lymphoma and neuroblastoma cells by RT-PCR. The presence of transcription factors in the different model cell lines was further investigated by EMSA analysis.
In pre-B cells mb-1 and CD19 are regulate by EBF-1 in collaboration with Pax-5 and E-proteins. We here show that neuroblastoma cells express these three, for B cell development crucial transcription factors, but nevertheless fail to express detectable levels of their known target genes. Expression of mb-1 could, however, be induced in neuroblastoma cells after disruption of the chromatin structure by treatment with 5-azacytidine and Trichostatin A.
These data suggest that transcription factors are able to selectively activate target genes in different tissues and that chromatin structure plays a key role in the regulation of this activity.
KeywordsNeuroblastoma Cell Neuroblastoma Cell Line Activate Target Gene Nalm6 Cell Tissue Specific Target Gene
The complex process of tumor development often involves changes in the transcription regulatory networks. In human cancer, genetic changes involving the transcription factor p53 gene is particularly common and the gene is found mutated in cancers originating from numerous cell types. This factor is, however, broadly expressed and directly involved in cell cycle regulation and apoptosis explaining the common involvement of the protein in tumor development. Many malignancies are characterized by specific chromosomal translocations that frequently affects the expression or structure of transcription factors with a more tissue specific expression pattern, often with an important function during development . Examples of this can be found within the hematopoetic system where for instance translocations of Tal-1 is associated with T cell leukemia's  while modified BCL-6  or c-myc  is associated with B cell non-Hodgkin's lymphomas. The close correlation with a specific tumor type and pathology to a specific transcription factor modification could well be explained by differential expression patterns and accessibility of the gene for translocation events. Another possibility could be that the action of the transcription factor is context dependent and therefore the ability of the modified protein to contribute to tumor development depends on the cell in which it arises.
To investigate mechanisms involved in lineage specific gene regulation and transcription factor target gene selection in tumor cells we have compared transcription factor expression in neuroblastoma and pre-B lymphoma cells. This revealed that both these highly divergent tumor types expressed the transcription factor EBF  that has been contributed a central role in B cell development . The protein is a helix-loop-helix family member [7, 8] essential for B-lymphopoesis in mice  where it has been shown to regulate a large number of pre-B cell restricted genes including the surrogate light chains , CD19  and the signal transduction proteins Igα (mb-1) [7, 11] and Igβ (B29) . The EBF-1 protein is highly conserved between human and mouse  and it also appears as if the target gene spectra has been conserved between species even though the primary promoter sequences of these genes has diverged [13, 14]. Pre-B cells express exclusively EBF-1  while neuroblastoma cells express other family members including EBF-2 and -3 . There are limited information about EBF target genes in neuroblastoma cells but binding sites for EBF proteins were identified in the promoters controlling the expression of the neuron restricted Chromogranin A (CGA) and SCG10 genes .
In the present study we were interested in the mechanisms of tissue specific target gene activation by comparing transcription factor function and gene expression patterns in neuroblastomas and pre-B cells. We here report that even though neuroblastoma cells express both Pax-5 and E-proteins, both suggested to be crucial co-activators for EBF target genes in pre-B cells, the cells do not express the pre-B cell restricted target genes mb-1 and CD19. The expression of the mb-1 gene could, however, be activated by treatment of the neuroblastoma cells with chromatin disrupting agents. This suggests that chromatin structure is a key component in the regulation of transcription factor function, by restricting the accessibility of target genes, possibly contributing to the apparent link between a transcription factor and specific malignancies.
HeLa, THP-1, KM3 and Nalm6 cells were grown in RPMI 1640 medium supplemented with 7.5% fetal calf serum (FCS), 10 mM HEPES, 2 mM pyruvate and 50 μg/ml gentamicin (complete RPMI media) (Life Technologies) at 37°C and 5% CO2. SH-SY5Y, IMR-32 and SK-N-BE(2)c neuroblastoma cells were cultured in Eagle's Minimum Essential Medium (MEM) with 10% FCS, 100 U/ml penicillin and 100 μg/ml streptomycin in an atmosphere of 5% CO2 at 37°C (Life Technologies). KCN-69n and LA-N-1 neuroblastoma cells were cultured in RPMI 1640 medium supplemented with FCS and antibiotics as above. The SK-N-BE(2)c cells were differentiated with 10 μM all-trans-retinoic acid (RA) in MEM containing 10% FCS for 0, 2, 8, 24, and 96 hours. 5-azacytidine (5-azaC) treatments were performed on SK-N-BE(2)c and SH-SY5Y cells at a concentration of 400 nM. After 64 h, fresh medium containing 50 nM Trichostatin A (TSA) was added. The cells were harvested in PBS after 24 h.
RNA was prepared from cells using Trizol (Life Technology) and cDNA was generated by annealing 1 μg of total RNA to 0,5 μg of random hexamers in 10 μl DEPC-treated water. Reverse transcriptase reactions were performed with 200 units of SuperScript Reverse Transcriptase (Life Technologies) in the manufacturers' buffer supplemented with 0,5 mM dNTP, 10 mM DTT and 20 units RNase inhibitor (Boeringer Mannheim, Bromma, Sweden) in a total volume of 20 μl, at 37°C for 1 hr. One-twentieth of the RT reactions were used in the PCR assays. PCR reactions were performed with 1 unit of Taq-polymerase (Life Technologies) in the manufacturers' buffer supplemented with 0.2 mM dNTP, in a total volume of 25 μl. GADPH was amplified by 25 cycles (94°C, 30 s, 55°C, 30 s and 72°C, 30 s) while 30 cycles were used to amplify CGA, SCG10, mb-1 and CD19 message (94°C, 30 s, 61°C, 30 s, 72°C 30 s). Primers were added to a final concentration of 1 mM. PCR products were blotted onto Hybond N+ nylon membranes (Amersham) using capillary blotting with 0.4 M NaOH. Membranes were pre-hybridized in 5X Denhardt's, 6XSSC, 0.1% SDS and 50 μg/ml Salmon Sperm DNA, at 57°C for 90 minutes and hybridized with γ[32P] labeled oligonucleotide for 12 hours at 57°C in the same solution. Membranes were washed at room temperature 2 times in 2XSSC supplemented with 0.1% SDS for 15 minutes.
Oligonucleotides used for RT-PCR were:
GADPH sense; 5'-CCACCCATGGCAAATTCCATGGCA;
GADPH antisense; 5'-TCTAGACGGCAGGTCAGGTCCACC;
CGA sense: 5'-GAAGATGAACTCTCAGAGGTTC
CGA antisense: 5'-GGATCTCCTTGTAGCCAAGGCTCG
CD19 sense: 5'-AGTCATTGCTGAGCCTAGAGCTG
CD19 antisense: 5'-CTCGGAGTCCTCCTCACTGTCAG
mb-1 sense; 5'-CCAGCATCATTGATGGTGAGCC
mb-1 antisense: 5'-GACATCTCCTATGTTGAGGCTGC
mb-1 hybridization; 5'-CCCGCACAATAGCAGCAACAACGCCAACGT
SCG10 sense; 5'-ATGCTGTCACTGATCTGCTCTTGC
SCG10 antisense; 5'-CAGGTTGAACTGTCTGGCTGAAG
DNA probes were labeled with γ[32P] ATP by incubation with T4 polynucleotide kinase (Roche Molecular Biochemicals), annealed and purified on a 5% polyacrylamide Tris-borate-EDTA (TBE) gel. Nuclear extract , or in vitro transcribed-translated protein, was incubated with labeled probe (20,000 cpm, 3 fmol) for 30 min at room temperature in binding buffer (10 mM HEPES pH [7.9], 70 mM KCl, 1 mM Dithiothreitol, 1 mM EDTA, 2.5 mM MgCl2, 1 mM ZnCl2, 5% Glycerol) with 0.75 μg Poly(dI/dC) (Amersham Pharmacia). Antibodies (anti Pax-5 SC-1974, anti Pu.1 SC-352, anti actin SC-1616, ets1/2 SC-275 and anti myc SC-764 all from Santa Cruz Biotech and anti E2-2 from Pharmingen) were added 10 min before the addition of the DNA probe. The samples were separated on a 6% polyacrylamide TBE gel, which was dried and subjected to autoradiography.
Oligonucleotides used for EMSA were the following:
mb-1 sense 5'-AGCCACCTCTCAGGGGAATTGTGG;
mb-1 antisense 5'-CCACAATTCCCCTGAGAGGTGGCT;
CD19-BSAP sense 5'-GCAGACACCCATGGTTGAGTGCCCTCCAGG;
CD19-BSAP antisense 5'-CCTGGAGGGCACTCAACCATGGGTGTCTGC;
μE5 sense: 5'-GGCCAGAACACCTGCAGACG;
μE5 antisense: 5'-CGTCTGCAGGTGTTCTGGCC;
Oct binding site sense 5'-CATCTCAAGTGATTTGCATCGCATGAGACG;
Oct binding site antisense 5'-CGTCTCATGCGATGCAAATCACTTGAGATC;
Lambda B (Pu.1 site) sense: 5'-GAAAAAGAGAAATAAAAGGAAGTGAAACCA AG;
Lambda B antisense: 5'-CTTGGTTTCACTTCCTTTTATTTCTCTTTTTC;
CRE (ATF5 site) sense: TCA TGG TAA AAA TGA CGT CAT GGT AAT TA
CRE antisense: TAA TTA CCA TGA CGT CAT TTT TAC CAT GA
cDNA micro-array analysis
RNA from Nalm6, SK-N-BE(2)c, SH-SY5Y, KM3, THP-1 and HeLa cells was extracted using Trizol™ (Invitrogen, Carlsbad, California). A common RNA control obtained by mixing a variety of cell lines was used for all hybridizations. RNA was concentrated to 50 μg total RNA (25 μg sample RNA and 25 μg control RNA) to generate aminoallyl-modified cDNA. Sample cDNA was labeled with Cy3-dCTP and control RNA was labeled with Cy5-dCTP using CyScribe Post-Labeling Kit (Amersham Pharmacia Biosciences). A hybridization solution was made by combining labeled cDNA with 20 μl Cot-1 DNA (1 mg/ml), 3 μl Poly dA (4 mg/ml) and 1.5 μl yeast t-RNA (4 mg/ml), dry down by speed-vac and resuspended in 40 μl Pronto! Universal Hybridization Solution™ (Pronto!™ Universal Microarray Reagent System, Corning). The hybridization solution was added to a pre-hybridized microarray slide (DNA microarrays were obtained from the SWEGENE DNA Microarray Resource Center, Lund University). The arrays were hybridized at 42°C for 18 hrs, washed according to the manufacturers recommendations (Pronto!™ Universal Microarray Reagent System, Corning), dried by centrifugation and scanned on Agilent microarray scanner. Scans were analyzed using GenePix Pro versions 22.214.171.124 and 126.96.36.199. BioArray Software Environment (BASE) (Saal et al. Genome Biology 2002, 3(8):software0003.1–0003.6). The settings for the analysis presented were, Background Correlation: Mean FG – Mean BG, Spot filter: (Raw) SNR ch1 mean > = 2, (Raw) SNR ch2 mean > = 2, (Raw) Flags = 0, (Raw) Spot diameter > = 40, Normalization: Lowess, Reporter filter: in # of assays = 16, Analysis: Hierarchical clustering (reporter)
Neuroblastoma and lymphoma cell lines display similarities in overall gene expression patterns but not of known EBF target genes
Neuroblastoma and pre-B lymphoma cells share the expression of the transcription factor Pax-5
Pre-B lymphoma and neuroblastoma cell lines display differential expression of E-, Ets and ATF proteins
To further investigate the E-box binding activities in SH-SY5Y cells as compared to Nalm6 cells we performed super-shift assays using antibodies against E2-2, E47 or c-myc (Figure 5B). The major complexes formed by nuclear extracts from the pre-B cell lines could be super-shifted either by the addition of anti E2-2 or anti E47 antibody while the complex formed in SH-SY5Y extracts reacted only on the addition of anti E2-2 antisera. None of the complexes in any of the cell lines were affected by the addition of either the c-myc or the actin antibody. This indicates that even though E2-2 is a major component in the formation of μE5 binding activity in the pre-B cells, it appears to form complex together with E47. This was not seen in the SH-SY5Y cells suggesting that there exist a distinct difference in the composition of E-box binding complexes in the two cell types.
EBF has also been suggested to share pre-B cell restricted target genes with Ets proteins and one factor belonging to this family of proteins suggested to contribute to B lineage identity is the Ets protein Pu.1 [28, 29]. In order to investigate the DNA binding activities in neuroblastoma and pre-B lymphoma we incubated a PU.1 binding site (λB) from the mouse Immunoglobulin λ enhancer with nuclear extracts from the two cell types (Figure 7). This showed a complex pattern of DNA binding activities but only the pre-B cells contained a complex reacting to the addition of anti Pu.1 anti-sera. Thus, we conclude, that even though neuroblastoma and lymphoma co-express several potent transcription factors there are also specific differences. Thus, even though both cell types express EBF, the protein would have to act in a different transcription factor context in the two cell types.
Chromatin structure participates directly in the regulation of tissue specific gene expression patterns
We here report that neuroblastoma and pre-B lymphoma cells share the expression of transcription factors believed to be essential for normal B cell development while the expression of their B lineage target genes remains tissue specific. The finding that neuroblastoma cells expressed Pax-5 was somewhat surprising even though this factor has been shown to be involved in midbrain development. Mice carrying a homologous disruption of the Pax-5 gene lack both B-lymphoid cells and a fully developed central nervous system [30, 31]. In the B cell compartment it also appears as if Pax-5 is crucial to lock the cell in the B-lymphoid developmental pathway. That is, even though pro-B cells from Pax-5 deficient mice express a whole set of early B cell markers [30, 31], they can, in contrast to normal pro-B cells, be differentiated into other hematopoetic lineages [32, 33]. Thus, Pax-5 appears to be essential for lineage fidelity in addition to be crucial for the completion of the B lymphoid development pathway. The former has been attributed to an ability of Pax-5 to inhibit the expression of cytokine receptors like M-CSF  and also of Notch-1 . Interestingly, Notch signaling plays a key role in fate selection of neuronal cells and we have previously shown that over expression of a constitutively active form of Notch-1 inhibits induced differentiation of neuroblastoma cells . It is currently not clear whether a link between Pax-5 and Notch1 expression is at hand in neuronal cells, but it is noteworthy that neuroblastoma cells express relatively high levels of Pax-5. Elevated expression levels of Pax-5 has been detected in two other tumors of neuronal origin, medulloblastoma  and astrocytoma  and in vitro data  as well as observed translocations in human B lineage tumors [38, 39, 40] indicate that Pax-5 have oncogenic properties.
In contrast to the stable expression of Pax-5 there appeared to be large differences in E-protein expression in the neuroblastoma cell lines investigated. In neuronal tissues a number of tissue specific bHLH proteins have been defined. Importantly, these pro-neuronal bHLH proteins require heterodimerization with E-proteins for the formation of DNA binding complexes. In the developing sympathetic nervous system HASH-1 is transiently expressed and is of pivotal importance for the formation of the autonomic and olfactory nervous system, neuroendocrine cells of the lung and specific regions of the telencephalon (reviewed in ). We have previously shown that a majority of the neuroblastomas express HASH-1, supporting the notion that the tumor is of embryonal origin . We also showed that E2-2 is the preferential binding partner of the pro-neuronal bHLH protein HASH-1 in SH-SY5Y neuroblastoma cells , and the EMSA results presented in this paper, showing that E2-2 is the main E-protein binding a μE5-E-box, corroborate this observation. Some findings indicate that E2-2 has specific and important functions in neuronal tissues [43, 44]. The gene is expressed at substantial level in brain compared to most other tissues  and E2-2 has been shown to bind and regulate several neuronal/neuroendocrine promoters, such as brain-specific FGF-1.B . Several lines of investigation has suggested that E-proteins are redundant in B cell development and that the apparent need of E2A protein in B cell development is due to lack of sufficient doses of functional E-proteins [43, 45]. This notion is substantiated by the finding that E2A deficiency can be rescued by expression of another E-protein, Heb , and that all these proteins appear able to activate the pre-B cell restricted λ5 promoter in synergy with EBF . Thus, based on these findings it is unlikely that the E-protein composition would have any dramatic impact on EBFs ability to activate target genes in neuroblastoma.
Our data also suggest that the chromatin structure of DNA is able to modulate the function of transcription factors in a dramatic manner. In the case of the mouse mb-1 promoter is has been shown that methylation of specific C residues in the promoter prevents the complex formation between Pax-5 and Ets proteins, thereby reducing functional activity of the promoter . Both the Ets and Pax-5 binding sites are conserved between mouse and human and similar mechanisms acting on the human promoter may well account for the observations in this report . It is also noteworthy that we obtained a much higher level of mb-1 transcription in the SH-SY5Y cells then in the SK-N-BE(2)c cells (Figure 7), a finding that could be expected due to that the SH-SY5Y cells expressed high levels of BSAP and E-proteins while SK-N-BE(2)c cells appeared to express somewhat lower levels of BSAP and no detectable amounts of μE5 binding E-proteins (Figure 3A and 5A). However, we were unable to observe any activation of CD19 gene even after treatment with demethylating agents suggesting that gene regulation is exerted at several different levels or that the human CD19 gene is highly dependent also of other factors, not present in the neuroblastoma cells. In this study we disrupt chromatin structure by treating the cells with a combination of a methylation inhibitor and a HDAC inhibitor. The processes of DNA methylation and histone acetylation are intimately connected. It has been known for a long time that histone deacetylase inhibitors, by reactivating gene expression, can inhibit growth and/or survival of cancer cells. Even though the precise mechanisms behind the effect of these drugs are largely unknown several of them are now evaluated in clinical trials. Furthermore, the observations of abnormal DNA methylation patterns in malignant cells are becoming increasingly interesting, as new information of the link between gene activation and methylation status increase.
One dilemma, however, resides in the issue if a gene is demethylated due to activation or activated due to demethylation. In our specific case the latter would be the most likely explanation. We cannot however exclude that our observations are due to the induction of other unknown proteins. The selective activation of mb-1 and the presence of all the known crucial transcription factors in the neuroblastoma cells would, however, be in line with de-methylation being the primary event. The shared expression of a whole set of transcription factors in such diverse malignancies as lymphoma and neuroblastoma may also be of importance for our view on the relationships between different tumor types. That is, even though these tumors are highly divergent and show large differences in the clinical outcome, there may be genetic links that can be used for drug targeting thus reducing the complexity of cancer treatment dramatically.
These data suggest that transcription factors are able to selectively activate target genes in different tissues and that chromatin structure plays a key role in the regulation of this activity.
We are grateful to the Swegene micro-array facility for help with cDNA expression analysis.
This work was supported by grants from the Swedish Cancer Society, the Children's Cancer Foundation of Sweden, The Swedish Research Council, Kock's Foundation and The Crafoord Foundation.
- 11.Sigvardsson M, Clark DR, Fitzsimmons D, Doyle M, Akerblad P, Breslin T, Bilke S, Li R, Yeamans C, Zhang G, Hagman J: Early B-cell factor, E2A, and Pax-5 cooperate to activate the early B cell-specific mb-1 promoter. Mol Cell Biol. 2002, 22: 8539-8551. 10.1128/MCB.22.24.8539-8551.2002.CrossRefPubMedPubMedCentralGoogle Scholar
- 15.Akerblad P, Lind U, Liberg D, Bamberg K, Sigvardsson M: Early B-cell Factor (O/E-1) is a promoter of adipogenisis and involved in control of genes important for terminal adipocyte differentiation. Mol Cell Biol. 2002, 22: 8015-8025. 10.1128/MCB.22.22.8015-8025.2002.CrossRefPubMedPubMedCentralGoogle Scholar
- 43.Liu Y, Ray SK, Yang XQ, Luntz-Leybman V, Chiu IM: A splice variant of E2-2 basic helix-loop-helix protein represses the brain-specific fibroblast growth factor 1 promoter through the binding to an imperfect E-box. J Biol Chem. 1998, 273: 19269-19276. 10.1074/jbc.273.30.19269.CrossRefPubMedGoogle Scholar
- The pre-publication history for this paper can be accessed here:http://www.biomedcentral.com/1471-2407/4/80/prepub
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.